Volcanic Lithology Logging Identification Based on ADASYN-KNN-RF Ensemble Model: Taking the Carboniferous System on the Hanging Wall of Kebai Fault Zone as an Example

Abstract The identification of volcanic lithology is crucial for the exploration and development of oil and gas reservoirs. Current methods primarily rely on core and thin-section data, which, despite their accuracy, are often inefficient, costly, and subjective. To address these limitations, conventional logging curves calibrated with core and thin-section data are employed in this study. A random forest intelligent identification model is established, and a novel ADASYN-KNN-RF ensemble model is further proposed for volcanic lithology identification. The sample numbers of various lithologies are first balanced using the ADASYN oversampling algorithm. Subsequently, both KNN and Random Forest (RF) models are trained independently. An optimal classification strategy is then determined by evaluating the performance of the two models on different lithologies, and their results are effectively integrated. When applied to actual data, the proposed ensemble model achieves a high identification accuracy of 93%. Comparative analysis demonstrates that the ensemble model outperforms individual models and proves to be more suitable for lithology identification.